My testresults:

unpatched ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

     nand read 0x20000000 0x00100000 0x03000000
         no problem

     nand write 0x20000000 0x00100000 0x00c00000
         no problem

     nand write 0x20000000 0x00100000 0x00d00000
         the watchdog will reset the processor

patched ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

     nand read 0x20000000 0x00100000 0x03000000
         no problem

     nand write 0x20000000 0x00100000 0x07f00000
         no problem

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

So, reading was never a problem, only writing.
And that is fixed now!
Thanks!


On 02/03/2011 01:23 AM, Scott Wood wrote:
> Poke the watchdog in a variety of looping constructs, which could take
> a long time to complete.
>
> Signed-off-by: Scott Wood<scottw...@freescale.com>
> ---
> Jaap, does this resolve the watchdog problems you were seeing?
>
>   drivers/mtd/nand/nand_base.c |    6 ++++++
>   drivers/mtd/nand/nand_util.c |    2 ++
>   2 files changed, 8 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/mtd/nand/nand_base.c b/drivers/mtd/nand/nand_base.c
> index 70c0593..b9bd394 100644
> --- a/drivers/mtd/nand/nand_base.c
> +++ b/drivers/mtd/nand/nand_base.c
> @@ -1156,6 +1156,8 @@ static int nand_do_read_ops(struct mtd_info *mtd, 
> loff_t from,
>       oob = ops->oobbuf;
>
>       while(1) {
> +             WATCHDOG_RESET();
> +
>               bytes = min(mtd->writesize - col, readlen);
>               aligned = (bytes == mtd->writesize);
>
> @@ -1485,6 +1487,7 @@ static int nand_do_read_oob(struct mtd_info *mtd, 
> loff_t from,
>       page = realpage&  chip->pagemask;
>
>       while(1) {
> +             WATCHDOG_RESET();
>               sndcmd = chip->ecc.read_oob(mtd, chip, page, sndcmd);
>
>               len = min(len, readlen);
> @@ -1884,6 +1887,8 @@ static int nand_do_write_ops(struct mtd_info *mtd, 
> loff_t to,
>               memset(chip->oob_poi, 0xff, mtd->oobsize);
>
>       while(1) {
> +             WATCHDOG_RESET();
> +
>               int bytes = mtd->writesize;
>               int cached = writelen>  bytes&&  page != blockmask;
>               uint8_t *wbuf = buf;
> @@ -2215,6 +2220,7 @@ int nand_erase_nand(struct mtd_info *mtd, struct 
> erase_info *instr,
>       instr->state = MTD_ERASING;
>
>       while (len) {
> +             WATCHDOG_RESET();
>               /*
>                * heck if we have a bad block, we do not erase bad blocks !
>                */
> diff --git a/drivers/mtd/nand/nand_util.c b/drivers/mtd/nand/nand_util.c
> index 8b4f738..5a6f7ae 100644
> --- a/drivers/mtd/nand/nand_util.c
> +++ b/drivers/mtd/nand/nand_util.c
> @@ -542,6 +542,8 @@ int nand_write_skip_bad(nand_info_t *nand, loff_t offset, 
> size_t *length,
>
>                       pages = write_size / pagesize_oob;
>                       for (page = 0; page<  pages; page++) {
> +                             WATCHDOG_RESET();
> +
>                               ops.datbuf = p_buffer;
>                               ops.oobbuf = ops.datbuf + pagesize;
>
_______________________________________________
U-Boot mailing list
U-Boot@lists.denx.de
http://lists.denx.de/mailman/listinfo/u-boot

Reply via email to